Bis(α-diimine)iron complexes: Electronic structure determination by spectroscopy and broken symmetry density functional theoretical calculations

Article abstract of DOI:10.1021/ic7022693

The electronic structure of a family comprising tetrahedral (α-diimine)iron dichloride, and tetrahedral bis(α-diimine)iron compounds has been investigated by Moessbauer spectroscopy, magnetic susceptibility measurements, and X-ray crystallography. In addition, broken-symmetry density functional theoretical (B3LYP) calculations have been performed. A detailed understanding of the electronic structure of these complexes has been obtained. A paramagnetic (S_t = 2), tetrahedral complex [Fe^II(⁴L)₂], where (⁴L) ^1- represents the diamagnetic monoanion N-tert-butylquinolinylamide, has been synthesized and characterized to serve as a benchmark for a Werner-type complex containing a tetrahedral Fe^IIN₄ geometry and a single high-spin ferrous ion. In contrast to the most commonly used description of the electronic structure of bis(α-diimine)iron(0) complexes as low-valent iron(0) species with two neutral α-diimine ligands, it is established here that they are, in fact, complexes containing two (α-diiminato)^1-? π radical monoanions and a high-spin ferrous ion (in tetrahedral N₄ geometry) (S_Fe = 2). Intramolecular antiferromagnetic coupling between the π radical ligands (S_rad = 1/2) and the ferrous ion (S_Fe = 2) yields the observed S_t = 1 ground state. The study confirms that α-diimines are redox noninnocent ligands with an energetically low-lying antibonding π* lowest unoccupied molecular orbital which can accept one or two electrons from a transition metal ion. The (α-diimine)FeCl ₂ complexes (S_t = 2) are shown to contain a neutral α-diimine ligand, a high spin ferrous ion, and two chloride ligands.

Full text of DOI:10.1021/ic7022693

Inorg. Chem. 2008, 47, 4579-4590
Bis(r-diimine)iron Complexes: Electronic Structure Determination by
Spectroscopy and Broken Symmetry Density Functional Theoretical
Calculations
Nicoleta Muresan,† Connie C. Lu,† Meenakshi Ghosh,† Jonas C. Peters,‡ Megumi Abe,‡
Lawrence M. Henling,‡ Thomas Weyhermöller,† Eckhard Bill,† and Karl Wieghardt*,†
Max-Planck Institute for Bioinorganic Chemistry, Stiftstrasse 34-36,
D-45470 Mu¨lheim an der Ruhr, Germany, and DiVision of Chemistry and Chemical Engineering,
Arnold and Mabel Beckman Laboratories of Chemical Synthesis, California Institute of
Technology, Pasadena, California 91125
Received November 19, 2007
The electronic structure of a family comprising tetrahedral (R-diimine)iron dichloride, and tetrahedral bis(R-diimine)iron
compounds has been investigated by Mo¨ssbauer spectroscopy, magnetic susceptibility measurements, and X-ray
crystallography. In addition, broken-symmetry density functional theoretical (B3LYP) calculations have been performed.
A detailed understanding of the electronic structure of these complexes has been obtained. A paramagnetic (St ) 2),
tetrahedral complex [FeII(4L)2], where (4L)1- represents the diamagnetic monoanion N-tert-butylquinolinylamide, has been
synthesized and characterized to serve as a benchmark for a Werner-type complex containing a tetrahedral FeIIN4 geometry
and a single high-spin ferrous ion. In contrast to the most commonly used description of the electronic structure of
bis(R-diimine)iron(0) complexes as low-valent iron(0) species with two neutral R-diimine ligands, it is established here
that they are, in fact, complexes containing two (R-diiminato)1-• π radical monoanions and a high-spin ferrous ion (in
1
tetrahedral N4 geometry) (SFe ) 2). Intramolecular antiferromagnetic coupling between the π radical ligands (Srad ) /2)
and the ferrous ion (SFe ) 2) yields the observed St ) 1 ground state. The study confirms that R-diimines are redox
noninnocent ligands with an energetically low-lying antibonding π* lowest unoccupied molecular orbital which can accept
one or two electrons from a transition metal ion. The (R-diimine)FeCl2 complexes (St ) 2) are shown to contain a neutral
R-diimine ligand, a high spin ferrous ion, and two chloride ligands.
Introduction
In 1977 tom Dieck and Bruder1 reported the synthesis of
an interesting series of bis(R-diimine)iron complexes which
they prepared by the reaction of FeCl2 with 2 equiv of an
R-diimine ligand under reducing conditions (2 equiv of
sodium) in diethylether. The electronic structure of these
violet, monomeric, putatively tetrahedral, and paramagnetic
(S ) 1) complexes [FeL2] has ever since1–4 been described
1
radical monoanions (Srad ) /2) which are intramolecularly
spin coupled yielding the observed triplet ground state, but
as zerovalent iron species (d8, S ) 1 in a tetrahedral ligand
field) containing two neutral R-diimine ligands.
We note that tom Dieck et al.2 indicated the possibility of
they did not provide any spectroscopic or crystallographic
evidence.
a more complex electronic structure involving a divalent,
high-spin ferrous ion (SFe ) 2) N,N′-coordinated to two π
In 2005 Chirik et al.4a reported the first crystal structure
of such bis(R-diimine)iron species. The quality of this
* To whom correspondence should be addressed. E-mail: wieghardt@
mpi-muelheim.mpg.de.
(1) tom Dieck, H.; Bruder, H. J. Chem. Soc., Chem. Commun. 1977, 24.
(2) tom Dieck, H.; Diercks, R.; Stamp, L.; Bruder, H.; Schuld, T. Chem.
Ber. 1987, 120, 1943.
†
Max-Planck-Institute for Bioinorganic Chemistry.
‡
California Institute of Technology.
10.1021/ic7022693 CCC: $40.75  2008 American Chemical Society
Inorganic Chemistry, Vol. 47, No. 11, 2008 4579
Published on Web 04/29/2008

Muresan et al.
Scheme 1
structure determination does not allow a detailed discussion
of C-N, C-C, and Fe-N distances. Nevertheless, these
authors concluded from an analysis of the structural data of
this and similar species4a,b that a description of the electronic
structure as intermediate spin ferrous species (SFe ) 1) with
“an important contribution” from the diamagnetic “ene-
diamide” and the neutral diamagnetic R-diimine canonical
form is appropriate.
Scheme 2
It is worth mentioning that a fourth electronic structure
should also be considered which involves an intermediate
spin ferric ion (SFe ) 3/2), a monoanionic π-radical, and a
diamagnetic dianion.
ions (xL•)1- (x ) 1,2,3) and the diamagnetic dianions
(xLRed)2-; a single resonance structure of each form is shown
in Scheme 2.
It is well established from careful X-ray crystallographic
studies of Zn-complexes that the ligand oxidation level differs
clearly in their respective C-N bond lengths (and also the
C-C bond distances).8–10
Recently, it has been shown that the neutral complexes of
some tetrahedral bis(R-diimine)nickel complexes possess an
electronic structure which is best described as nickel(II) (d8,
SNi ) 1 in tetrahedral ligand field) with two N,N′-coordinated
monoanionic ligand π radicals (L•)1- yielding a diamagnetic
ground state.11,12
Finally, we have reinvestigated some of Chirik’s complexes,
namely those shown in Scheme 3 (7, 8, and 9). We have
recorded their zero-field Mo¨ssbauer spectra and conclude that
Bis(R-diimine)iron complexes are important species be-
cause they are known to initiate catalytic atom-transfer
radical polymerization,5 the dimerization of dienes,6 and the
polymerization of olefins.7
We decided to synthesize a new series of such complexes
using the R-diimine ligands shown in Scheme 1, namely
(1LOx)0, (2LOx)0, and (3LOx)0.
The ligand abbreviations were chosen to highlight the fact
that neutral R-diimines may be reduced in two single one-
electron steps generating the paramagnetic π radical monoan-
(6) (a) tom Dieck, H.; Stamp, L.; Diercks, R.; Mu¨ller, C. New J. Chem.
1985, 9, 289. (b) Le Floch, P.; Knoch, F.; Kremer, F.; Mathey, F.;
Scholz, J.; Scholz, W.; Thiele, K.-H.; Zenneck, U. Eur. J. Inorg.
Chem.m 1998, 119.
(3) Walther, D.; Kreisel, G.; Kirmse, R. Z. Anorg. Allg. Chem. 1982, 487,
149.
(7) Lorber, C; Choukroun, R.; Costes, J.-P.; Donnadien, B. C. C. R. Chim.
2002, 5, 251.
(4) (a) Bart, S. C.; Hawrelak, E. J.; Lobkovsky, E.; Chirik, P. J.
Organometallics 2005, 24, 5518. (b) Bart, S. C.; Hawrelak, E. J.;
Schmisseur, A. K.; Lobkovsky, E.; Chirik, P. J. Organometallics 2004,
23, 237.
(8) van Koten, G.; Vrieze, K. AdV. Organomet. Chem. 1982, 21, 151.
(9) Gardiner, M. G.; Hanson, G. R.; Henderson, M. J.; Lee, F. C.; Raston,
C. L. Inorg. Chem. 1994, 33, 2456.
(5) (a) Gibson, V. C.; O’Reilly, R. K.; Wass, D. F.; White, A. J. P.;
Williams, D. J. Macromolecules 2003, 36, 2591. (b) Gibson, V. C.;
O’Reilly, R. K.; Reed, W.; Wass, D. F.; White, A. J. P.; Williams,
P. J. Chem. Commun. 2002, 1850.
(10) Rijnberg, E.; Richter, B.; Thiele, K.-H.; Boersma, J.; Veldman, N.;
Spek, A. L.; van Koten, G. Inorg. Chem. 1998, 37, 56.
(11) Khusniyarov, M. M.; Harms, K.; Burghaus, O.; Sundermeyer, J. Eur.
J. Inorg. Chem. 2006, 15, 2985.
4580 Inorganic Chemistry, Vol. 47, No. 11, 2008

Bis(r-diimine)iron Complexes
Scheme 3
of nBuLi, red-orange lithium tert-butylquinolinylamide precipitated
from the reaction. The reaction was warmed up to 0 °C, and the
solvent removed in vacuo to yield a red solid. In the dry box, the
red solid was washed with petroleum ether (3 × 20 mL) to yield
a bright orange powder (1.83 g, 94%). Protonation with H2O and
analysis by GC-MS showed only N-tert-butyl-quinolinylamine.
Synthesis of Complexes. [FeII(4L)2] (1). Lithium N-tert-bu-
tylquinolinylamide (1.2 g, 5.82 mmol) dissolved in 20 mL of tetra-
hydrofuran was added to a stirring suspension of FeCl2 (368.8 mg,
2.91 mmol) in 30 mL of tetrahydrofuran. After stirring at room
temperature overnight, a dark red-brown homogeneous solution was
obtained. Tetrahydrofuran was removed in vacuo to yield dark red
solids. The dark red solids were washed with petroleum ether and
stirred in 25 mL of toluene overnight. The toluene solution was
filtered over a frit, and dark red solids were collected. Solids were
washed with toluene (4 × 20 mL) until only a gray powder
remained. Toluene was removed in vacuo to yield a dark red-brown
powder. The powder was then washed with 10% benzene in
petroleum ether (4 × 5 mL), 20% benzene in petroleum ether (4
× 5 mL), and 30% benzene in petroleum ether (1 × 5 mL), and
dried in vacuo (0.819 g, 62%). 1H NMR (C6D6, 300 MHz): δ 64.02
(s); 61.21 (bs); 52.88 (s); -7.42 (s); -51.31 (s); -55.35 (s). Evans
Method (C6D6, 298 K): 4.8 µB Anal. Calcd for C26H30N4Fe: C, 68.7;
H, 6.6; N, 12.3. Found: C, 68.9; H, 6.6; N, 12.5.
FeII(1LOx)Cl2] (2). To a solution of 2-phenyl-1,4-bis(isopropyl)-
1,4-diazabutadiene (0.5 g, 2.31 mmol) in tetrahydrofuran (20 mL)
under an argon blanketing atmosphere was added 0.29 g of FeCl2
(2.31 mmol). The deep violet solution was stirred for 20 h at room
temperature. The solvent was removed by evaporation under
reduced pressure, and the residue was washed with toluene.
Complex 2 was isolated as a dark violet solid (0.75 g, 2.18 mmol,
94%). X-ray quality crystals were obtained by slow evaporation of
the solvent from a concentrated solution of 2 in tetrahydrofuran.
Anal. Calcd for C14H20N2FeCl2: C, 49.1; H, 5.8; N, 8.2; Fe, 16.3;
Cl, 20.7. Found: C, 49.1; H, 6.0; N, 7.9; Fe, 16.1; Cl, 20.5.
[FeII(1L)2] (3). To a solution of 2 (0.6 mg, 1.75 mmol) in n-hexane
(30 mL) was added sodium (80 mg, 3.50 mmol) with stirring at 20
°C. After 10 min., a solution of 2-phenyl-1,4-bis(isopropyl)-1,4-
diazabutadiene (0.37 g, 1.75 mmol) in n-hexane (10 mL) was added,
and the reaction mixture was stirred for 24 h. The dark violet color of
the solution turned to dark brown. The solvent was removed by
evaporation under reduced pressure to give a dark brown precipitate
which was washed with acetonitrile and dried in vacuo. Yield: 0.76 g
(89%). Anal. Calcd for C28H40N4Fe: C, 68.8; H, 8.2; N, 11.5; Fe, 11.4.
Found: C, 68.8; H, 8.5; N, 11.3; Fe, 11.1.
none of these contain a low-valent iron ion with a formal
oxidation number of 0 or +1. Instead, all of them are identified
as high-spin ferrous complexes (d6, SFe ) 2) with zero (7), one
(9), and two (8) ligand π radical monoanionic R-diimine ligands
per metal. Note that 9 is a dimer with two µ-chloro bridges,
and intramolecular antiferromagnetic exchange coupling is
anticipated. Therefore, we have investigated the temperature
dependence of its magnetic properties.
Experimental Section
All air-sensitive materials were manipulated under a N2 or Ar
blanketing atmosphere by using standard Schlenk line procedures
or a glovebox. All commercially available chemicals were used as
received. The ligands 2-phenyl-1,4-bis(isopropyl)-1,4-diazabutadi-
ene (1LOx)0, 2-methyl-1,4-bis(2,6-diisopropylphenyl)-1,4-diaza-1,3-
butadiene (2LOx), and enantiopure R-(-)-bis(phenylimino)camphene
(3LOx)0 were synthesized according to published procedures.13a–c
Chirik’s complexes 7, 8, and 9 have been synthesized as described
in refs 4a and 4b.
Synthesis of N-tert-Butylquinolinylamine, H(4L). Pd(OAc)2
(151.1 mg, 0.629 mmol) and BINAP (419 mg, 0.629 mmol) were
preheated at 85 °C for 30 min in 15 mL of toluene. The mixture
was added to a 250 mL reaction bomb flask containing bromo-
quinoline (7.0 g, 33.64 mmol), tert-butylamine (3.691 g, 50.47
mmol), and sodium tert-butoxide (4.527 g, 47.1 mmol) in 75 mL
of toluene. The resulting reaction mixture was dark purple. The
reaction bomb was sealed with a Teflon cap and heated at 110 °C
for 48 h. The solution was then cooled to room temperature in air,
diluted with 200 mL of dichloromethane, and washed with H2O (3
× 200 mL) and brine (2 × 150 mL). The organic layer was
collected, dried over MgSO4, and filtered over celite. Removal of
solvent in vacuo yielded a dark brown oil, which was purified by
flashing through a column of silica with dichloromethane as the
eluant. Dichloromethane is removed in vacuo to yield a bright
orange-yellow oil (6.318 g, 94%). 1H NMR (CDCl3, 300 MHz): δ
8.69 (dd, J1 ) 4.0 Hz, J2 ) 1.6 Hz, 1H); 8.03 (dd, J1 ) 8.0 Hz, J2
) 1.6 Hz, 1H); 7.37-7.32 (m, 2H); 7.00 (t, J1 ) 8.0 Hz, J2 ) 4.0
Hz, 1H); 6.94 (dd, J1 ) 8.0 Hz, J2 ) 4.0 Hz, 1H); 6.45 (s, 1H,
NH); 1.54 (s, 9H, t-butyl); GC-MS: m/z ) 200.
[FeII(2LOx)Cl2] (4). To a solution of 2-methyl-1,4-bis(2,6-diiso-
propylphenyl)-1,4-diaza-1,3-butadiene (0.5 g, 1.79 mmol) in tetra-
hydrofuran (20 mL) under an argon blanketing atmosphere was
added 0.22 g of FeCl2 (1.79 mmol). The deep blue solution was
stirred for 20 h at room temperature. The solvent was removed by
evaporation under reduced pressure, and the residue was washed
with toluene to remove any soluble impurities. Complex 4 was
isolated as a brown solid (0.70 g, 1.72 mmol, 97%). Anal. Calcd
Preparation of Lithium N-tert-Butylquinolinylamide, Li(4L). 1.6 M
nBuLi (5.93 mL, 9.49 mmol) was added dropwise to a stirring
solution of tert-butylquinolinylamine (2.00 g, 9.99 mmol) in 75
mL of petroleum ether at -78 °C under nitrogen. Upon addition
(13) (a) Armesto, D.; Bosch, P.; Gallego, M. G.; Martin, J. F.; Ortiz, M. J.;
Perez-Ossorio, R.; Ramos, A. Org. Prep. Proced. Int. 1987, 19, 181.
(b) Druzhkov, N. O.; Teplova, I. A.; Glushakova, V. N.; Skorodumova,
N. A.; Abakumov, G. A.; Zhao, B.; Berluche, E.; Schulz, D. N.; Baugh,
L. S.; Ballinger, C. A.; Squire, K. R.; Canich, J.-A. M.; Bubnov,
M. P. O. S. Patent: PCT/U.S. Patent 022356, 2003; WO Patent 007509,
2004. (c) van Asselt, R.; Elsevier, C. J.; Smeets, W. J. J.; Spek, A. L.;
Benedix, R Rec. TraV. Chim. (Pays Bas) 1994, 113, 88.
(12) (a) Muresan, N.; Chlopek, K.; Weyhermu¨ller, T.; Neese, F.; Wieghardt,
K. Inorg. Chem. 2007, 46, 5327. (b) Blanchard, S.; Neese, F.; Bothe,
E.; Bill, E.; Weyhermu¨ller, T.; Wieghardt, K. Inorg. Chem. 2005, 44,
3636. (c) Chlopek, K.; Bothe, E.; Neese, F.; Weyhermu¨ller, T.;
Wieghardt, K. Inorg. Chem. 2006, 45, 6298.
Inorganic Chemistry, Vol. 47, No. 11, 2008 4581

Muresan et al.
Table 1. Crystallographic Data for 1, 2, and 6
for C19H22N2FeCl2: C, 56.3; H, 5.4; N, 6.9; Fe, 13.8. Found: C,
56.1; H, 5.3; N, 7.0; Fe, 13.8.
1
2
6
[FeII(2L)2] (5). To a solution of 4 (0.6 mg, 1.48 mmol) in
n-hexane (30 mL) was added sodium (68 mg, 2.96 mmol) with
stirring at 20 °C. After 10 min, a solution of 2-methyl-1,4-bis(2,6-
diisopropylphenyl)-1,4-diaza-1,3-butadiene (0.41 g, 1.48 mmol) in
n-hexane (10 mL) was added, and the reaction mixture was stirred
for 24 h. The dark blue color of the solution turned to dark green.
The solvent was removed by evaporation under reduced pressure
to give a dark green precipitate which was washed with acetonitrile
and dried in vacuo. Yield: 0.76 g (80%). Anal. Calcd for C38H44-
N4Fe: C, 74.53; H, 7.24; N, 9.15; Fe, 9.12. Found: C, 74.0; H,
7.30; N, 9.15; Fe, 9.20.
[FeII(3L•)2] (6). Two equiv of the ligand (1R)-(-)-(phenylimi-
no)camphane (778 mg, 2.45 mmol), 1 equiv of FeCl2 (159 mg,
1.23 mmol), and 2.05 equiv of sodium metal (58.0 mg, 2.52 mmol)
were added to a glass vessel. Dimethoxyether (DME, 10 mL) was
added, and the reaction mixture was vigorously stirred for 18 h.
The solvent was then completely removed under reduced pressure.
Pentane (2 mL) was added to the crude residue, which was then
dried further under reduced pressure. The dried residue was washed
with pentane (2 × 3 mL), extracted with C6H6, filtered through
celite, and subjected to reduced pressure for solvent removal. A
dark-brown powder was thus obtained. Single crystals of X-ray
quality were grown from vapor diffusion of pentane into a
concentrated C6H6 solution of 6. Yield: 711 mg (84%). Anal. Calcd
for C44H48FeN4: C, 76.73; H, 7.02; N, 8.13. Found: C, 76.6; H,
7.1; N, 8.2.
chem. formula
fw
space group
a, Å
b, Å
c, Å
C26H30FeN4
454.39
C14H20Cl2FeN2 C44H48FeN4
343.07
688.71
j
C2/c, No. 15 P1, No. 2
P21, No. 4
14.1374(3)
13.7276(3)
19.2999(5)
90
96.481(3)
90
3721.6(2)
4
17.8934(14)
14.2722(11)
17.8740(14)
90
91.372(1)
90
4563.3(6)
8
98(2)
7.4721(3)
9.9945(5)
11.9345(5)
71.501(3)
78.244(3)
87.333(3)
827.33(6)
2
R, deg
ꢀ, deg
γ, deg
V, Å
Z
T, K
100(2)
1.377
120(2)
F calcd, g cm-3
refl.collected/2Θmax
unique refl./I > 2σ(I)
no. of params/restr.
λ, Å/µ(KR), cm-1
R1a/goodness of fitb
wR2c (I > 2σ(I))
1.323
1.229
33432/56.90 22997/70.00
110059/67.00
29054/26787
1029/139
5356/4372
287/0
7265/6505
176/0
0.71073/6.81 0.71073/12.23 0.71073/4.41
0.0361/2.259 0.0291/1.023
0.0689 0.0676
0.0512/1.045
0.1262
+0.43/-0.54
residual density, e Å-3 +0.87/-0.48 +0.58/-0.39
a Observation criterion: I > 2σ(I). R1 ) Σ||Fo|-|Fc||/Σ|Fo|. b GOF )
c wR2 ) [Σ[w(Fo - Fc )2]/Σ[w(Fo )2]]1/2
2
2
2
2
2
[Σ[w(Fo - Fc )2]/(n - p)]1/2
.
2
2
2
where w ) 1/σ2(Fo ) + (aP)2 + bP, P ) (Fo + 2Fc )/3.
and therefore, the ligand substructures within each disordered
position were restrained to be equal within errors using the SAME
instruction of ShelXL97. In addition, equal anisotropic displacement
parameters were used for corresponding atoms giving a total of
139 restraints.
X-ray Crystallography. X-Ray Crystallographic Data Collec-
tion and Refinement of the Structures. A dark red single crystal
of 1, a pink colored crystal of 2, and a brown specimen of 6 were
coated with perfluoropolyether, picked up with nylon loops and
mounted in the nitrogen cold stream of the diffractometers. A
Bruker-Nonius KappaCCD diffractometer equipped with a Mo-
target rotating-anode X-ray source was used in case of 2 and 6,
and a Bruker Smart 1000 CCD diffractometer with a sealed tube
source was used for compound 1. Graphite monochromated Mo
KR radiation (λ )0.71073 Å) was used throughout. Final cell
constants were obtained from least-squares fits of all measured
reflections. Intensity data were corrected for absorption using
intensities of redundant reflections. The structures were readily
solved by Patterson methods and subsequent difference Fourier
techniques. The Siemens ShelXTL14 software package was used
for solution and artwork of the structures; ShelXL9715 was used
for the refinement. All non-hydrogen atoms were anisotropically
refined, and hydrogen atoms were placed at calculated positions
and refined as riding atoms with isotropic displacement parameters.
Crystallographic data of the compounds are listed in Table 1.
Compound 6 crystallizes in the chiral space-group P21, and the
unit cell was found to contain two crystallographically independent
molecules. Refinement of the structure revealed that cocrystalliza-
tion of the two possible enantiomerically pure geometrical isomers
(cisoid and transoid) occurred for both positions, but only one of
the two ligands in each independent molecule is statically disor-
dered. The occupation factors refined to a ratio of about 50% for
the trans- and the cis-isomer. The absolute structure parameter
refined to a value of 0.036 (0.008), and the obtained structure
matches nicely the absolute configuration of the employed ligand.
It was assumed that the metrical parameters of the ligand are not
significantly influenced by the coordination mode (cis or trans),
Physical Measurements. Magnetic susceptibility data were
measured from a powder sample in the temperature range 2-300
K by using a SQUID susceptometer (MPMS-7, Quantum Design)
with a field of 1.0 T. The experimental data were corrected for
underlying diamagnetism by use of tabulated Pascal’s constants,
as well as for temperature-independent paramagnetism. The sus-
ceptibility data of 9 were simulated with our own program julX
for exchange coupled systems (see Supporting Information). The
simulations are based on the spin-Hamiltonian operator for a linear
tetranuclear system of two high-spin Fe(II) with spin S2 ) 2, S3 )
1
1
2, and two terminal ligand radicals with S1 ) /2, S4 ) /2.
ˆ
ˆ
ˆ
ˆ
ˆ
ˆ
ˆ
2 3
H ) -2J S · S + S · S - 2J ′ S · S
[
]
4
1
2
3
b
b
ˆ
ˆ
ˆ
ˆ
4
+ gꢀ S · S · B + gꢀ S + S · B
(
)
3
(
)
2
1
2
2
2
ˆ
ˆ
ˆ
+
D S - (1 ⁄ 3)S(S + 1) + (E ⁄ D) S - Si,y
(1)
[
(
)]
∑
i,z
i,x
i)2,3
Here J and J′ are the exchange coupling constants for the iron-
radical and the iron-iron coupling, respectively; g and g′ are the
average electronic g values for iron and radicals, and D and E/D
are the axial zero-field splitting and rhombicity parameters for iron.
The magnetic moments were obtained from the first order derivative
of the eigenvalues of eq 1.
The equipment used for the IR, UV-vis, and Mo¨ssbauer
spectroscopies has been described previously (Ghosh, P.; Bill, E.;
Weyhermu¨ller, T.; Wieghardt, K. J. Am. Chem. Soc. 2003, 125,
3967 and Ray, K.; Bill, E.; Weyhermu¨ller, T.; Wieghardt, K. J. Am.
Chem. Soc. 2005, 127, 5641).
Calculations. All calculations were performed by using the
ORCA program package.16 The geometry optimizations were
(16) Neese, F. ORCA, an Ab Initio, Density Functional and Semiempirical
Electronic Structure Program Package, Version 2.4, Revision 36;
Max-Planck-Institut fu¨r Bioanorganische Chemie: Mu¨lheim/Ruhr,
Germany, May 2005.
(14) ShelXTL Version 6.14; Bruker AXS Inc.: Madison, WI, 2003.
(15) Sheldrick, G. M. ShelXL97, University of Go¨ttingen: Go¨ttingen,
Germany, 1997.
4582 Inorganic Chemistry, Vol. 47, No. 11, 2008

Bis(r-diimine)iron Complexes
carried out at the B3LYP level17 of density-functional theory (DFT).
The all-electron Gaussian basis sets were those reported by the
Ahlrichs group.18,19 Triple-ꢁ quality basis sets with one set of
polarization functions on the iron and nitrogen atoms were used
(TZVP).19 The carbon and hydrogen atoms were described by
slightly smaller polarized split-valence SV(P) basis sets that are
double-ꢁ quality in the valence region and contain a polarizing set
of d-functions on the non-hydrogen atoms.18 The auxiliary basis
sets for all complexes used to expand the electron density in the
calculations were chosen to match the orbital basis. The self-
consistent field (SCF) calculations were tightly converged (1 × 10-8
Eh in energy, 1 × 10-7 Eh in the density change, and 1 × 10-7 in
the maximum element of the DIIS error vector). The geometries
were considered converged after the energy change was less than
5 × 10-6 Eh, the gradient norm and maximum gradient element
were smaller than 1 × 10-4 Eh/Bohr and 3 × 10-4 Eh/Bohr,
respectively, and the root-mean square and maximum displacements
of atoms were smaller than 2 × 10-3 Bohr and 4 × 10-3 Bohr,
respectively. Corresponding orbitals20 and density plots were
obtained by the program Molekel.21
Nonrelativistic single point calculations on the optimized ge-
ometries of iron complexes with the B3LYP functional were carried
out to predict Mo¨ssbauer spectral parameters (isomer shift and
quadrupole splitting). These calculations employed the CP(PPP)
basis set22 for iron and the TZV(P) basis sets for N atoms. The
SV(P) basis sets were used for the remaining atoms. The Mo¨ssbauer
isomer shifts were calculated from the computed electron densities
at the iron centers as previously described.23
Throughout this paper we describe our computational results of
iron complexes containing noninnocent ligands using the broken
symmetry (BS) approach.24–26 Recently, the BS approach was
successfully applied for transition-metal-containing complexes in
the groups of Wieghardt and Neese27–33 and others.25,34
fragment 2. In most cases fragments 1 and 2 correspond to the
metal and the ligand, respectively. Note that in this notation a
standard high-spin open-shell solution would be written down as
BS(m+n,0). In general, The BS(m,n) notation refers to the initial
guess to the wave function. The variational process does, however,
have the freedom to converge to a solution of the form BS(m-n,0)
where effectively the n-spin-down electrons pair with n < m spin-
up electrons on the partner fragment. Such a solution is then a
standard MS ) (m - n)/2 unrestricted Kohn-Sham solution. As
explained elsewhere,20 the nature of the solution is investigated
via the corresponding orbital transformation, which via the corre-
sponding orbital overlaps displays whether the system is to be
described as a spin-coupled or a closed-shell solution.
Results and Discussion
Synthesis of Complexes. Scheme 1 summarizes the
ligands and the new complexes 1-6, and Scheme 3 shows
the ligands and the revisited complexes 7-9 reported by
Chirik and co-workers.4
The reaction of FeCl2 in dry tetrahydrofuran with 2 equiv
of lithium tert-butylquinolinylamide (Li(4L), Scheme 1)
affords a dark red solid of [FeII(4L)2] (1) upon evaporation
of the solvent. The effective magnetic moment of 1 has been
determined by NMR spectroscopy (Evans method in C6D6)
to be 4.5 µB at 298 K which indicates the expected high
spin S ) 2 ground-state for 1. As shown by X-ray
crystallography, the high-spin ferrous ion is in a nearly
regular tetrahedral environment of 4 nitrogen donor atoms.
The two bidentate ligands are the redox-innocent, closed
shell, monoanion tert-butylquinolinylamide(1-), (4L)1-.
Similarly, the reaction of FeCl2 in dry tetrahydrofuran
under strictly anaerobic conditions with 1 equiv of an
R-diimine, (1LOx)0 or (2LOx) (Scheme 1), yields the neutral,
dark-violet, tetrahedral complexes [FeII(1LOx)Cl2] (2) and
[FeII(2LOx)Cl2] (4), respectively. Temperature dependent
(3-300 K) magnetic susceptibility measurements of solid 2
and 4 in an 1.0 T magnetic field shown in Figure 1 for 4
reveal again an S ) 2 ground state (high spin ferrous ions).
Table 2 summarizes the spin Hamiltonian parameters.
Using the same conditions as tom Dieck and Bruder1
described for the synthesis of their neutral bis(R-diimine)iron
complexes, namely, the reaction of 2 equiv of the ligand
(1LOx)0, (2LOx)0, and enantiopure (3LOx)0 with 1 equiv of
FeCl2and 2 equiv of sodium in n-hexane and DME respectively,
under an argon atmosphere, the following brown-green com-
plexes were obtained in good yields: [FeII(1L•)2] (3), [FeII(2L•)2]
(5), and [FeII(3L•)2] (6). It is also possible to generate 3 and 5
from 2 and 4 by addition of 2 equiv of (1LOx) and (2LOx),
respectively, and 2 equiv of sodium in n-hexane.
Because for the complexes studied in this work, one can obtain
broken-symmetry (BS) solutions to the spin-unrestricted Kohn-Sham
equations, we will adopt the following notation: the system is
divided into two fragments. The notation BS(m,n) refers to a broken
symmetry state with m unpaired spin-up electrons on fragment 1
and n unpaired spin-down electrons essentially localized on
(17) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (b) Becke, A. D.
J. Chem. Phys. 1986, 84, 4524. (c) Lee, C. T.; Yang, W. T.; Parr,
R. G. Phys. ReV. B 1988, 37, 785.
(18) Scha¨fer, A.; Horn, H.; Ahlrichs, R. J. Chem. Phys. 1992, 97, 2571.
(19) Scha¨fer, A.; Huber, C.; Ahlrichs, R. J. Chem. Phys. 1994, 100, 5829.
(20) Neese, F. J. Phys. Chem. Solids 2004, 65, 781.
(21) Molekel, AdVanced InteractiVe 3D-Graphics for Molecular Sciences;
available under http://www.cscs.ch/molekel/.
(22) Neese, F. Inorg. Chim. Acta 2002, 337, 181.
(23) Sinnecker, S.; Slep, L. D.; Bill, E.; Neese, F. Inorg. Chem. 2005, 44,
2245.
(24) Noodleman, L.; Norman, J. G. J. Chem. Phys. 1979, 70, 4903.
(25) Noodleman, L.; Peng, C. Y.; Case, D. A.; Mouesca, J. M. Coord.
Chem. ReV. 1995, 144, 199.
(26) (a) Noodleman, L.; Case, D. A.; Aizman, A. J. Am. Chem. Soc. 1988,
110, 1001. (b) Noodleman, L.; Davidson, E. R. Chem. Phys. 1986,
109, 131. (c) Noodleman, L.; Norman, J. G.; Osborne, J. H.; Aizman,
A.; Case, D. A. J. Am. Chem. Soc. 1985, 107, 3418. (d) Noodleman,
L. J. Chem. Phys. 1981, 74, 5737.
As shown in Figure 2 for 5, VTVH magnetic susceptibility
measurements revealed that 3, 5, and 6 possess an S ) 1
(27) (a) Bachler, V.; Olbrich, G.; Neese, F.; Wieghardt, K. Inorg. Chem.
2002, 41, 4179. (b) Ghosh, P.; Bill, E.; Weyhermu¨ller, T.; Neese, F.;
Wieghardt, K. J. Am. Chem. Soc. 2003, 125, 1293.
(31) Chlopek, K.; Bothe, E.; Neese, F.; Weyhermu¨ller, T.; Wieghardt, K.
Inorg. Chem. 2006, 45, 6298.
(28) Blanchard, S.; Neese, F.; Bothe, E.; Bill, E.; Weyhermu¨ller, T.;
Wieghardt, K. Inorg. Chem. 2005, 44, 3636.
(32) Patra. A. K.; Bill, E.; Bothe, E.; Chlopek, K.; Neese, F.; Weyhermu¨ller,
T.; Stobie, K.; Ward, M. D.; McCleverty, J. A.; Wieghardt, K. Inorg.
Chem. 2006, 45, 7877.
(29) (a) Herebian, D.; Bothe, E.; Neese, F.; Weyhermu¨ller, T.; Wieghardt,
K. J. Am. Chem. Soc. 2003, 125, 9116. (b) Herebian, D.; Wieghardt,
K.; Neese, F. J. Am. Chem. Soc. 2003, 125, 10997.
(33) (a) Remenyi, C.; Kaupp, M. J. Am. Chem. Soc. 2005, 127, 11399. (b)
Bencini, A.; Carbonera, C.; Dei, A.; Vaz, M. G. F. Dalton Trans.
2003, 1701.
(30) Slep, L. D.; Mijovilovich, A.; Meyer-Klaucke, W.; Weyhermu¨ller, T.;
Bill, E.; Bothe, E.; Neese, F.; Wieghardt, K. J. Am. Chem. Soc. 2003,
125, 15554.
(34) Baker, R. J.; Farley, R. D.; Jones, C.; Mills, D. P.; Kloth, M.; Murphy,
D. M. Chem.sEur. J. 2005, 11, 2972.
Inorganic Chemistry, Vol. 47, No. 11, 2008 4583

Muresan et al.
Figure 1. Temperature dependent magnetic moments, µeff/µB, and a variable
temperature, variable field magnetization plot (inset) of [FeII(2LOx)Cl2] (4).
(Simulation parameters are given in Table 2).
Table 2. Spin Hamiltonian Parameters from Variable Temperature,
Variable Field (VTVH) Magnetic Susceptibility Measurements (1 T)
b
c
Sta
g
|Dt|, cm-1
|DFe|, cm-1
1
2
3
4
5
6
2
2
1
2
1
1
2.0
2.01
2.06(fixed)
2.02
2.05(fixed)
2.30
n.m.d
7.7
18.5
7.2
16.1
16.2
7.7
8.1
7.2
7.7
8.1
a Ground state; b Zero-field splitting of St state; c Zero-field splitting
parameter of intrinsic SFe state obtained by using spin projection techniques.
d n.m. ) not measured.
Figure 3. Electronic spectra of 3 (top) and 5 (bottom) in tetrahydrofuran.
Table 3. Electronic Spectra of Complexes in Tetrahydrofuran Solution
at 20 °C
complex
λ, nm (ε, 104 M-1 cm-1
)
1 (S ) 2)
2 (S ) 2)
3 (S ) 1)
367(0.55), 495(0.8), 940(0.02)
363(0.16), 551(0.03), 610(0.03)
331(0.5), 441(0.6), 515(0.2), 560(0.15), 951(0.016),
1100sh (0.012)
4 (S ) 2)
5 (S ) 1)
323(0.22), 359(0.20), 584(0.03), 648(0.04)
326(3.55), 392(2.53), 456(1.55), 573(0.39), 694(0.3),
1317(0.006)
6 (S ) 1)
392(0.8), 482sh(0.3), 550(0.2), 715(0.04), 990s(0.02),
1130sh(0.02)
Figure 2. Temperature dependent magnetic moments, µeff/µB, and a VTVH
magnetization plot (inset) of [FeII(2L•)2] (5). (Simulation parameters are
given in Table 2). Similar plots of 3 and 6 are given in the Supporting
Information.
projection techniques. They agree nicely with values for 2
and 4. Thus, complexes 2-6 all contain a high spin ferrous
ion in a tetrahedral ligand field.
Electronic spectra of complexes in tetrahydrofuran solution
have been recorded at 20 °C; the results are summarized in
Table 3. Figure 3 displays the spectrum of 3 (and 5) which
exhibits a weak ligand-to-ligand charge transfer band (LLCT)
at ∼1100 nm (ε ) 120 M-1 cm-1). This band is also present
in the corresponding tetrahedral [Ni(L•)2] complexes.12a,35,36
Complex 5 displays this transition at 1317 nm (ε ) 60 M-1
cm-1). This LLCT band is spin-forbidden in tetrahedral
[FeII(L•)2] complexes and is therefore weak.35 Similarly, in
complex 6 this LLCT band has been identified at 1130 nm.
This transition is absent in the spectra of 1, 2, and 4 all of
which contain closed shell ligands and a high spin ferrous
ground state; the spin Hamiltonian parameters are sum-
marized in Table 2. As we will show below, these neutral
tetrahedral complexes possess a central high spin ferrous ion
(SFe ) 2) and two N,N′-coordinated ligand π radical anions
(Srad ) 1/2); the spins couple intramolecularly antiferromag-
netically yielding the observed triplet ground state.
It is interesting that the zero-field splitting parameter |D|
values for the S ) 2 complexes 2 and 4 are at ∼7 cm-1
whereas the corresponding values |D|t for the S ) 1 species
3, 5, and 6 are larger at ∼17 cm-1. The latter values may be
converted to the intrinsic zero-field splitting parameter of
the central iron ion (SFe ) 2) |D|Fe ∼ 7 cm-1 by using spin
4584 Inorganic Chemistry, Vol. 47, No. 11, 2008

Bis(r-diimine)iron Complexes
ion (SFe ) 2). The LLCT band is therefore considered to be
a spectroscopic marker for the presence of two ligand π
radicals in [FeII(L•)2]0 species.
We have also prepared Chirik’s complexes 7, 8, and 9
(Scheme 3) according to the reported procedures4 to inves-
tigate more completely the electronic structure of these
complexes and recorded their Mo¨ssbauer spectra (see below).
These complexes contain the ArNdC(CH3)sC(CH3)dNAr
(Ar ) 2,6-diisopropylphenyl) ligand (5LOx). [FeII(5LOx)Cl2]
(7) is the exact analog of complexes 2 and 4 which all contain
a high spin ferrous ion and a neutral R-diimine ligand, as
well as two chloride ligands in a tetrahedral geometry. In
accord with this, the magnetic susceptibility data of 7 indicate
also an S ) 2 ground state.
Figure 4. Temperature-dependence of magnetic moments of 9 (experi-
mental data). The solid line represents a best fit of the data (see text).
The complex 8 containing two R-diimine type radical
monoanions (our interpretation as we will show below) and
a high spin ferrous ion (SFe ) 2) possesses a triplet ground
state (St ) 1). Note that the authors in ref 4a prefer to
describe its electronic structure as a tetrahedral iron(0) d8
complex containing two closed-shell R-diimine ligands or
as an intermediate-spin iron(III) d5 complex where one
R-diimine acts as a closed-shell, strong field, dianionic ligand
and the second as a monoanionic π radical. The reported
crystal structure determination of 8 is inconclusive and
appears to be flawed because the two C-N bonds within
each bidentate ligand are found to be inequivalent(!) at
1.35(1) and 1.43(1) Å (1.28(1) and 1.35(1) Å in the other
ligand).37
Table 4. Experimental and Calculated Bond Distances (Å) of 1
exp.
calcd.
Fe1-N1
Fe1-N2
N1-C1
N1-C9
N2-C8
N2-C10
C1-C2
C2-C3
C3-C4
C4-C5
C4-C9
C5-C6
C6-C7
C7-C8
C8-C9
θ, dega
2.083(1)
1.972(1)
1.331(2)
1.370(2)
1.366(2)
1.474(2)
1.394(2)
1.372(2)
1.413(2)
1.415(2)
1.417(2)
1.365(2)
1.403(2)
1.401(2)
1.452(2)
87
2.126
2.014
1.326
1.364
1.365
1.481
1.409
1.381
1.421
1.418
1.433
1.383
1.410
1.413
1.463
88.8
Finally, reduction of 7 with LiCH2CHMe2 yields the dimer
9 which has been characterized by X-ray crystallography.4b
Two tetrahedrons are bridged by two chloride ligands (edge-
sharing).
a Dihedral angle between two five-membered chelate rings.
the presence of a large number of well spread spin sublevels
in the corresponding energy range.
It has not been possible to simulate the data satisfactorily
3
by using a model where only two effective spins S* ) /2
(arising from a very strong antiferromagnetic exchange
coupling of a high spin ferrous ion coordinated to a ligand
radical (SFe ) 2 and Srad
)
1/2)) are intramolecularly
We have measured its temperature-dependent molar
magnetic susceptibility in a 1.0 T magnetic field. The
temperature-dependent magnetic moments of 9 in the range
2-295 K are shown in Figure 4; they display a continual
increase from 0.9 µB at 2 K to 5.4 µB at 295 K, revealing
antiferromagnetically exchange coupled. On the other hand,
if the ligand-to-metal interaction is not predominant, one can
explicitly consider this by using a model which involves a
spin tetramer L• · · · FeII · · · FeII · · · L• with two terminal ligand
1
radicals (5L•)1- (Srad ) /2) and two high spin ferrous ions
(SFe ) 2). By using the spin-Hamiltonian in eq 1 and
including a 9% paramagnetic impurity (S ) 5/2) as indicated
by the Mo¨ssbauer spectrum and a fixed zero-field splitting
for the iron(II) sites (|D| ) 12 ( 5 cm-1), the following
exchange coupling constants were obtained from a best fit:
J ) J1 ) J3 ) -53 cm-1 for the iron-radical interaction and
J′ ) J2 ) -25 cm-1 for the iron-iron coupling.
X-ray Crystal Structures. The structures of 1, 2, and 6
have been determined by single crystal X-ray crystallography
at 100(2) K. Crystallographic data are summarized in Table
1 and selected bond distances are given in Tables 4 and 5.
Figures 5, 6, and 7 show the structures of 1, 2, and 6,
respectively.
(35) Muresan, N.; Weyhermu¨ller, T.; Wieghardt, K. Dalton Trans. 2007,
4390.
(36) (a) tom Dieck, H.; Svoboda, M.; Greiser, Z. Z. Naturforsch. 1981,
36b, 823. (b) Svoboda, M.; tom Dieck, H.; Kru¨ger, C.; Tsay, Y.-H. Z.
Naturforsch. 1981, 36b, 814. (c) Bonrath, W.; Po¨rschke, K. R.; Mynott,
R.; Kru¨ger, C. Z. Naturforsch. 1990, 45b, 1647.
(37) No crystallographic details for 8 have been given in the text of ref 4a
(which is compound 4 in this paper). Retrieval from the supporting
information material and a closer inspection of the CIF-file, which
has been deposited with the Cambridge Crystallographic Data Center,
revealed that the compound has been refined in an incorrect space
group. Although the correct space group P2/n (No.13) is given in the
crystallographic table and the CIF-file, the compound was actually
refined in the non-centrosymmetric space-group Pn (No.7) and these
data were used to calculate bond distances. The missing crystal-
lographic C2 axis brings about the apparent inequivalence of the C-N
bond distances in the chelate rings and the suspicious thermal
displacement parameters.
Inorganic Chemistry, Vol. 47, No. 11, 2008 4585

Muresan et al.
Table 5. Calculated and Experimental Bond Distances (Å) of
Five-Membered Chelate Rings and Dihedral Angles θ (deg) between
Chelate Rings in FeL2 Complexes (L ) R-diimine)
3
5
[FeII(5′L•)2] transoid-6 transoid-6
(calcd) (calcd)
(calcd)
(exp.)
(calcd)
Fe-N1
2.038
2.006
2.028
2.014
1.348
1.420
1.336
1.350
1.420
1.335
2.108
2.088
2.103
2.090
1.345
1.412
1.338
1.346
1.413
1.337
2.058
2.049
2.059
2.049
1.347
1.431
1.347
1.344
1.431
1.349
79
2.013(2)
2.025(2)
disordered
disordered
1.343(2)
1.408(3)
1.341(2)
disordered
disordered
disordered
73.0
2.051
2.049
2.056
2.051
1.341
1.419
1.341
1.341
1.420
1.340
76.4
Fe-N4
Fe-N21
Fe-N24
N1-C2
C2-C3
C3-N4
N21-C22
C22-C23
C23-N24
a
θ, deg
88
56
a Dihedral angle between chelate rings.
Figure 7. Structure of the transoid form of 6 (that of the cisoid form is
not shown).
ligands are unexceptional and in the usual range observed
for quinolines. This indicates that the closed-shell monoan-
ions (4L)1- in 1 are redox-innocent. Thus, complex 1 serves
as a benchmark for a tetrahedral, high spin ferrous species
(S ) 2) with an FeN4 polyhedron.
Neutral molecules [FeII(1LOx)Cl2] in crystals of 2 contain
a four coordinate, high spin ferrous ion with a neutral N,N′-
coordinated R-diimine ligand (1LOx)0 and two chloride
ligands. The FeN2Cl2 polyhedron is distorted tetrahedral; the
two planes FeCl2 and FeN2 form a dihedral angle of 88.9°.
The average Fe-N and Fe-Cl bond lengths are long at 2.112
and 2.239 Å, respectively; they indicate a high spin config-
uration at the central FeII (S ) 2). Two very similar structures
have recently been reported by Gibson et al.38 It is now very
gratifying that the average C-N bond length of the ligand
(1LOx)0 at 1.284 Å in 2 corresponds to a CdN double bond,
whereas the distance C2sC3 at 1.501 Å is a typical single
bond. The oxidation level of the R-diimine ligand is therefore
that of neutral (1LOx)0 as shown in Schemes 1 and 2. Thus,
complex 2 serves as a benchmark for a closed shell, neutral
R-diimine ligand coordinated to a high spin ferrous ion.
Complex 6 crystallizes in the chiral monoclinic space
group P21 with two crystallographically independent neutral
molecules [FeII(3L•)2] in the unit cell. Interestingly, these
molecules display a static disorder comprising the transoid-
and cisoid-isomer which exist only if the dihedral angle θ
between the two five-membered chelate rings is * 90°.
In both independent molecules only one of the ligands is
disordered. Because a number of restraints had to be
introduced to refine a split atom model, metrical parameters
of the disordered parts are less reliable. In the following we
will therefore discuss the structural parameters of the
nondisordered part of the transoid-isomer only. The averaged
dihedral angle θ of the two independent molecules in 6 is
73.0°. The FeN4 polyhedron is distorted tetrahedral, and the
average Fe-N bond length at 2.02 Å is long and indicates
Figure 5. Structure of a neutral molecule in crystals of 1 (hydrogen atoms
were omitted; thermal ellipsoids are drawn at the 50% level).
Figure 6. Structure of a neutral molecule in crystals of 2 (thermal ellipsoids
are at the 50% level).
Neutral molecules [FeII(4L)2] in crystals of 1 contain two
N,N′-coordinated monoanionic N-tert-butylquinolinylamide
and a high spin ferrous ion. The FeN4 coordination polyhe-
dron is a tetrahedron where the two Namide-Fe-Npy planes
form a dihedral angle θ of 87° (it is 90° in a regular
tetrahedron and 0° in a square planar arrangement). The
average Fe-Namide and Fe-Npy distance at 1.986 and 2.083
Å, respectively, are long and consistent with four-coordinated
high spin FeII. The C-C and C-N bond lengths of the
(38) Allan, L. E. N.; Shaver, M. P.; White, A. J. P.; Gibson, V. C. Inorg.
Chem. 2007, 46, 8963.
4586 Inorganic Chemistry, Vol. 47, No. 11, 2008

Bis(r-diimine)iron Complexes
Table 6. Mo¨ssbauer Parameters of Complexes at 80 Ka
complex
c
d
e
Stb
δ, mm s-1
|∆EQ|, mm s-1
%
1
2
0.77(0.70)
0.48
2.14(2.15)
0.6
94.2
5.8
2
3
2
1
0.89
0.50(0.51)
0.43
2.78
4.28(4.45)
0.83
100
91.7
8.3
4
5
6
2
1
1
1.02
0.97
0.59(0.68)
0.43
0.65(0.51)
0.65
3.13
2.52
2.82(2.95)
0.83
4.19(4.43)
3.62
83
17
91.2
8.8
the presence of a high spin ferrous ion (SFe ) 2). The two
ligands are identical within the 3σ standard deviation limit
of the C-C and C-N bond distances. The oxidation level
of the two R-diimines is clearly that of a π radical monoanion
(3L•)1- as shown in Scheme 2 because the average C-N
distance at 1.342 Å is intermediate between an R-diimine
58 transoid
40 cisoid
a Values given in parentheses are calculated ones (B3LYP). b Ground
state. c Isomer shift vs R-Fe (298 K). d Quadrupole splitting. e Amount of
major and minor component (subspectra a and b).
Table 7. Zero-field Mo¨ssbauer Data of Complexes 7, 8, 9 at 80 Ka
complex
c
d
e
Stb
SFe
δ, mm s-1
|∆EQ|, mm s-1
(3LOx)0 at ∼1.29 Å and an enediamide (3LRed 2-
)
at 1.38 Å.
7
8
9
2
2
2
2
0.84
0.77(0.73)
0.88
2.57
1.81(2.08)
2.68
Similarly, the av C-C distance at 1.408 Å is intermediate
between that in (3LOx)0 and (3LRed)2-. We note that the quality
of the present structure determination of 6 is significantly
better than that of complex 8 in ref 4. Here the two C-N
distances in either of the two ligands are identical.
1
0f
a Values in parenthesis were calculated. b Ground-state of molecule.
c Intrinsic spin state of central iron ion. d Isomer shift vs R-Fe (298 K).
e Quadrupole splitting. f Antiferromagnetic coupling of two high spin ferrous
ions (see text).
An excellent crystal structure determination of complex
94 has been reported. This complex contains two edge-
sharing LFeCl2 tetrahedral units. Interestingly, the reported
C-C and C-N distances of the R-diimine ligand shown in
Figure 8 are in excellent agreement with the data for a π
radical anion (5L•)1- (Scheme 2). This implies that the central
metal ion possesses a +II oxidation level (high spin ferrous
as was deduced from the magnetic data).
Mo¨ssbauer Spectroscopy. Zero-field Mo¨ssbauer spectra
of polycrystalline samples were recorded at 80 K; the results
are summarized in Table 6 and 7 The spectra of 1-5 consist
of a doublet from a major component (>80%) and a minor
doublet from some unknown impurities (<20%). It is now
important that the spectra of the tetrahedral high spin species
1, 2, and 4 with an S ) 2 ground-state exhibit very similar
large isomer shift values δ in the range 0.77-1.02 mm s-1
and large quadrupole splitting parameters in the range
2.14-3.13 mm s-1. These values are characteristic for high
spin ferrous species (SFe ) 2) in a tetrahedral ligand field.
The tetrahedral complexes 3, 5, and 6 possess an S ) 1
ground state. The isomer shift values of these species are
observed in the narrow range 0.50-0.70 mm s-1; the quadru-
pole splitting parameters are very large in the range 2.82-4.28
mm s-1. This set of data points again to the presence of high
Figure 9. Zero-field Mo¨ssbauer spectrum of 6 at 80 K: Subspectrum a
(40%) may correspond to the cisoid form of the neutral molecule in 6
whereas that of b (58%) may be due to the corresponding transoid isomer.
spin ferrous ions (SFe ) 2). (We note that to the best of our
knowledge there are no authenticated Mo¨ssbauer parameters
for a genuine Fe(0) (d8) species with an S ) 1 spin state.)
Interestingly, as shown in Figure 9 the zero-field Mo¨ss-
bauer spectrum at 80 K of solid 6 consists of two very similar
quadrupole doublets, namely subspectrum a (∼40%) with δ
) 0.65 mm s-1 and ∆EQ ) 3.62 mm s-1 and subspectrum
b with δ ) 0.65 mm s-1 and ∆EQ ) 4.19 mm s-1 (∼58%).
A third subspectrum c (∼2.5%) with δ ) 0.55 mm s-1 and
∆EQ ) 0.53 mm s-1 is probably due to a high spin ferric
species produced via oxidatiave degradation of 6 with air.
Subspectrum a corresponds to the cisoid form of 6 whereas
that of b is probably the spectrum of the transoid form. This
would be in agreement with the X-ray crystallographic results
(vide supra).
Figure 10 displays magnetically perturbed Mo¨ssbauer
spectra of 3. For the simulations the following parameters
from magnetic susceptibility measurements were used (Table
2): gx ) gy ) gz ) 2.06; |D| ) 18.5 cm-1, St ) 1; E/D ) 0.
The results give δ ) 0.50 mm s-1, ∆EQ ) -4.28 mm s-1
(at 4.2 K); an asymmetry parameter η ) 0.2; A/gNꢀN(St), T
) (-11.0, -16.0, -32.0). By using spin projection tech-
niques it is possible to convert the magnetic hyperfine tensor
values A/gNꢀN (St ) 1) to the intrinsic values which are
Figure 8. Crystallographic data for complex 9 from ref 4. Bond distances
are given in Å.
Inorganic Chemistry, Vol. 47, No. 11, 2008 4587

Muresan et al.
Zero-field Mo¨ssbauer data for complexes 7, 8, and 9 are
summarized in Table 7. The high isomer shift values of all
three species clearly point to the presence of high spin ferrous
ions (SFe ) 2) in all three cases. This implies that 7 contains
a neutral R-diimine ligand [FeII(5LOx)Cl2] whereas 8 contains
two π radical monoanions as in [FeII(5L•)2]. Most interest-
ingly, the dimeric complex 9 must also contain two high
spin ferrous ions and two ligand π radical anions.
DFT Calculations. a. Optimized Geometries for 1, 3,
5, and [FeII(5′L•)2]. The tetrahedral complex 1 possesses
an S ) 2 ground state. We optimized its ground-state
geometry assuming the presence of a high spin, open shell
S ) 2 state by using the spin-unrestricted Kohn-Sham
functional B3LYP. The experimental and calculated
geometric structures are in excellent agreement (Table 4);
the C-C and C-N bond distances agree within (0.01 Å
whereas the calculated Fe-N bond distances are overes-
timated by ∼0.04 Å which is a typical result for the
B3LYP functional. It is also remarkable that the calculated
dihedralangleθbetweenthetwofive-memberedFe-N-C-
C-N chelate rings at 88.8° is in excellent agreement with
the experimental value of 87°.
Tetrahedral complexes 3, 5, 6, and [FeII(5′L•)2], the latter
of which is the truncated model of complex [FeII(5L•)2]
(Scheme 3)4 where the i-propyl groups have been replaced
by hydrogen atoms, possess a triplet ground state. Therefore,
we have calculated the following models: (a) a standard spin-
unrestricted S ) 1 model; (b) a broken symmetry BS(4,2)
MS ) 1 model for a high spin ferrous ion (SFe ) 2) and two
antiferromagnetically coupled ligand π radicals (Srad ) 1/2),
and (c) a BS(3,1) MS ) 1 model implying the presence of
3
an intermediate spin ferric ion (SFe ) /2), one ligand π
radical, (L•)1-, and one closed shell enediamide ligand
(LRed)2-. Interestingly, all three models for each complex
converged to the same isoenergetic and isostructural solution
of the BS(4,2) minimum. The adiabatically excited BS(6,0)
model was found to be ∼11 kcal mol higher in energy than
the corresponding BS(4,2) solution.
Table 5 summarizes the calculated metrical details of the two
five-membered chelate rings Fe-N-C-C-N in 3, 5, transoid-
6, and [FeII(5′L•)2]. It is remarkable that irrespective of the
N-substituents of the respective R-diimine or of the carbon
backbone, the C-N and C-C bond lengths of the five-membered
chelate rings are within a very small range identical with those
reported for the monoanionic π radical form in Scheme 2.
Experimentally this backbone has the same bond lengths in 6.
Therefore, we conclude that the structural data (calculated or
experimental) are in excellent agreement with the presence of two
monoanionic ligand π radicals (xL•)1- in these complexes.
We note that the dihedral angle θ between the two five-
membered chelate rings is strongly influenced by the steric bulk
of the N-substituents and varies from 88° in 3 to 56° in 5. The
smaller this substituent is, the more tetrahedral is the FeN4
polyhedron (θ f 90°).
Figure 10. Mo¨ssbauer spectra of 3 with B ) 7T applied perpendicular to
the γ-ray.
b. Electronic Structures of 1, 3, 5, transoid-6, and [FeII-
(5′L•)2]. For pseudotetrahedral 1, the molecular orbital (MO)
bonding scheme (Supporting Information) shows the pres-
characteristic for the local spin state of the ferrous ion (SFe
) 2): A/gNꢀN (SFe ) 2) ) (-7.3, -10.7, -21.3) T. These
values are quite typical for a high spin ferrous ion.
4588 Inorganic Chemistry, Vol. 47, No. 11, 2008

Bis(r-diimine)iron Complexes
Figure 12. Spin-density plot shown with values as derived from the BS(4,2)
calculation of 3 (based on a Mulliken spin population analysis).
two unpaired electrons with a negative spin localized on the
two ligand π radicals.
The corresponding orbital transformation is used to visual-
ize the overlapping magnetic pairs of the system.20 The spin-
orbitals obtained from single-point unrestricted calculations
were transformed in such a way that for each spin-up orbital
there exists at most one spin-down partner that has nonzero
spatial overlap. Values of S close to 1 indicate a standard
doubly occupied MO with little spin-polarization, whereas
S , 1 is the signature of nonorthogonal magnetic orbital
pairs. For 3 two such magnetically interacting pairs which
interact via a π pathway have been identified. Each of these
pairs consists of one metal orbital and the corresponding
ligand radical orbital.
The computational electronic structure results for
[FeII(5′L•)2] and transoid-6 (Supporting Information) give
exactly the same model as shown above for 3. Two ligand
π radicals are intramolecularly antiferromagnetically coupled
to a high spin ferrous ion yielding the observed triplet ground
state.
Figure 11. Qualitative MO diagram of the magnetic orbitals derived from
BS(4,2) calculation of 3. The spatial overlap (S) of corresponding alpha
and beta orbitals is given.
ence of a central (higher spin) ferrous ion: the five highest
energy orbitals are of predominantly metal-d character, one
of which is doubly occupied, and four are singly occupied
as one expects for an S ) 2 Fe(II) ion in a tetrahedral ligand
field. The Mulliken spin population analysis corroborates this
notion: ∼3.8 unpaired electrons are located at the ferrous
ion and no spin density resides on the ligands. Thus, 1 is a
classical Werner-type coordination compound.
A qualitative MO bonding scheme derived from the spin-
unrestricted BS(4,2) calculation for 3 is shown in Figure 11.
Five orbitals of predominantly metal-d character are again
identified. One of them is found in the spin-up and spin-
down manifolds; it is doubly occupied. The other four Fe-
based orbitals occur only in the spin-up manifold. These four
orbitals are thus singly occupied with parallel spins (Fe(II);
SFe ) 2). In addition, two ligand centered orbitals are
identified in the spin-down manifold which are not populated
in the spin-up manifold. This situation leads to the observed
overall MS ) 1 state. These orbitals correspond to two
symmetry-adapted combinations of the singly occupied
molecular orbital (SOMO) of the two ligand π radicals.
Complex 3 features a high spin Fe(II) ion which is antifer-
romagnetically coupled to two ligand-centered π radical
monoanions.
To calibrate the above calculated electronic structures, we
have calculated the Mo¨ssbauer spectra22,23 of selected
complexes, namely those of 1, 3, 5, transoid-6, and of
complex 8. The results are summarized in Table 6 and 7. It
is quite remarkable how well the experimental and calculated
values agree. For transoid-6 the calculated and experimental
geometric structure and Mo¨ssbauer parameters are in good
agreement giving confidence in the correct evaluation of the
electronic structures for 3, for 5 where X-ray structures are
not available, and for 8 where the crystal structure is
flawed.37 Thus, all present bis(R-diimine)iron complexes (3,
5, 6, and 8) with a triplet ground-state must be described as
high-spin ferrous species N,N′-coordinated to two π radical
monoanions (L•)1- where strong intramolecular antiferro-
magnetic spin-coupling yields the observed S ) 1 ground
state.
The spin density plot (Mulliken analysis) in Figure 12
nicely shows the antiparallel spin alignment between the high
spin Fe(II) (positive spin density in red) and the π radical
ligands (negative spin density in yellow). An approximate
breakdown of the spin density into atomic contributions via
a spin population analysis supports the presence of four
unpaired electrons at the iron ion with a positive spin and
Conclusion
From a combination of structural and spectroscopic
investigations in conjunction with broken-symmetry DFT
calculations it has been possible to unequivocally establish
the following electronic structures for a variety of iron
Inorganic Chemistry, Vol. 47, No. 11, 2008 4589

Muresan et al.
complexes containing R-diimine type ligands in two different
oxidation levels, namely neutral ligand (xLOx)0 and monoan-
ionic π radicals (xL•)1-.
Complex 7 is a previously described4 high-spin ferrous
complex (S ) 2) with an N,N′-coordinated neutral R-diimine
ligand and two coordinated chloride anions. Complex 8
is also a high-spin ferrous species but N,N′-coordinated
to two R-diiminate π radical anions. Finally, dimeric 9
contains two high-spin ferrous ions and two π radical
monoanionic ligands (5L•)1- yielding, via intramolecular
antiferromagnetic exchange coupling, the observed S )
0 ground state.
(1) The tetrahedral, neutral complex 1 (S ) 2) is a typical
Werner-type complex with a central high-spin ferrous ion
and two closed-shell, monoanionic N-tert-butylquinolinyla-
mide ligand. Four unpaired electrons are located in metal-d
orbitals, and no spin density has been detected on the ligands.
(2) The tetrahedral neutral complexes (S ) 2) 2, 4, and 7
also contain a central ferrous ion coordinated to two chloride
ions and one neutral, closed-shell R-diimine ligand (the
average C-N bond is found at 1.29 Å).
(3) The most salient feature of the present study is the
discovery that the neutral bis(R-diimine)iron complexes 3,
5, 6, and 8 with a triplet ground-state all consist of a high-
spin ferrous ion (SFe ) 2) N,N′-coordinated to two R-diimi-
nate(1-) π radicals (Srad ) 1/2). The observed triplet ground-
state is attained via intramolecular antiferromagnetic spin-spin
coupling between the ligand π radicals and the ferrous ion.
BS(4,2) DFT calculations clearly show (from a Mulliken spin
population analysis) R-spin density on both ligands (∼2
unpaired electrons) and ꢀ-spin density at the central metal
ion (4 unpaired electrons). There is no spectroscopic or DFT
calculational evidence for the description as low-valent Fe(0)
with two diamagnetic R-diimine ligands.
Acknowledgment. N.M. and M.G. are grateful to the
Max-Planck Society for a postdoctoral stipend and C.C.L.
thanks the A. v. Humboldt Foundation for a fellowship. We
gratefully acknowledge financial support from the Fonds of
the Chemical Industry. M.A. thanks the Caltech SURF
program for funding. We also would like to thank Dr. K.
Chlopek for preliminary DFT calculations of 8.
Supporting Information Available: X-ray crystallographic files
in CIF format for 1, 2, and 6 and tables of geometrical and electronic
structural details of BS DFT calculated structures (PDF). Figures
S1 and S2 display temperature dependent magnetic susceptibility
data of 2 and 6; Figures S3 and S4 show disorder models of 6.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(4) As a consequence of this study, the electronic structure
descriptions of complexes 7, 8, and 9 need to be revised.
IC7022693
4590 Inorganic Chemistry, Vol. 47, No. 11, 2008